| *** ekcs has quit IRC | 01:23 | |
| *** ekcs has joined #openstack-self-healing | 02:00 | |
| *** ekcs has quit IRC | 05:25 | |
| *** Vedanu has joined #openstack-self-healing | 06:59 | |
| Vedanu | Hi | 06:59 |
|---|---|---|
| Vedanu | Self healing component: | 07:00 |
| Vedanu | Below is the framework I am proposing.. please comment | 07:00 |
| Vedanu | 1.Defining a healing group | 07:00 |
| Vedanu | It is a group which indicates what are the resources that are needed for the service to be up like the VM/application. | 07:00 |
| Vedanu | Eg 1: if a horizon application to be running then apache service to be up and a corresponding IP Address should be pingable and the NIC should be operating | 07:00 |
| Vedanu | Eg 2: If a VM with the floating IP Address assigned, neutron service to be up, then the IP should be routable, NIC should be up | 07:01 |
| Vedanu | Eg 3: If a VM has disk storage attached, then the cinder service to be up, external storage device should be up, mount points should be reachable NIC/FC to be up. | 07:01 |
| Vedanu | 2.Creating a resource dependencies | 07:01 |
| Vedanu | It indicates that sequence of bringing up the resources. And also indicates the resource type | 07:01 |
| Vedanu | 3.Heartbeat indicator | 07:01 |
| Vedanu | Indicates which is the compute which has failure even after the re-tries and needs the evacuation | 07:01 |
| Vedanu | Zabbix can be configured to raise alarm for the failure and depending on the alarm a broadcast msg has to be sent to all the other computes to respond back with its resource availability | 07:01 |
| Vedanu | Eg: if an alarm is raised on the issues with the mount points then only the VM which has external disk associated has to be marked for the migration/evacuation | 07:01 |
| Vedanu | 4.Migration | 07:01 |
| Vedanu | VM’s marked for the evacuation is moved to the compute which are rich in its resources | 07:01 |
| *** bogdando has joined #openstack-self-healing | 07:04 | |
| *** alexchadin has joined #openstack-self-healing | 08:02 | |
| *** pcichy has joined #openstack-self-healing | 08:32 | |
| *** Vedanu has quit IRC | 08:49 | |
| *** alexchadin has quit IRC | 08:53 | |
| *** Vedanu has joined #openstack-self-healing | 09:01 | |
| *** alexchadin has joined #openstack-self-healing | 09:07 | |
| *** alexchadin has quit IRC | 09:55 | |
| *** Vedanu has quit IRC | 10:20 | |
| *** Vedanu has joined #openstack-self-healing | 10:45 | |
| *** alexchadin has joined #openstack-self-healing | 11:09 | |
| *** Vedanu has quit IRC | 11:10 | |
| *** Vedanu has joined #openstack-self-healing | 11:21 | |
| *** Vedanu has quit IRC | 11:52 | |
| * aspiers waits for Vedanu to rejoin | 12:49 | |
| aspiers | IRC really needs to die a horrible death | 12:49 |
| *** alexchadin has quit IRC | 13:02 | |
| *** alexchadin has joined #openstack-self-healing | 13:07 | |
| *** alexchadin has quit IRC | 13:33 | |
| *** alexchadin has joined #openstack-self-healing | 13:34 | |
| *** alexchadin has quit IRC | 13:45 | |
| *** pcichy has quit IRC | 13:49 | |
| *** bogdando has quit IRC | 17:17 | |
| *** ekcs has joined #openstack-self-healing | 17:32 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!